Vision Transformer Based Adaptive Beamforming for GNSS Bands

Global Navigation Satellite Systems (GNSS) are the most used navigation method nowadays. However, GNSS signals are often jammed on purpose. Therefore, it became necessary to eliminate these jammer signals to obtain navigation data. In this study, a vision transformer based adaptive beamforming method (ViT-BF) is presented to suppress jammer signals. As the difference from the previously studied ViT based beamforming approaches, the model is built with autocorrelation matrix as input and beamforming weights as output, which provides a blind beamforming method. ViT-BF approach is compared with a previously proposed convolutional neural network based beamforming (CNN-BF) and noise subspace tracking (NST), which ViT-BF is resulted more successfully in terms of beam and null divergences in directions of signal arrivals and with a shorter response time.